AI Assistant Raters (US Based)

You will review prompts and responses produced by an assigned AI assistant and judge their quality and usefulness in English. The goal is to assess how well responses serve everyday users.

You will use your own personal account on Gemini, Instinct, or Muse to support evaluations. Your account must have conversation history from more than two weeks of use, and you will apply consistent judgment across response-rating tasks.

• Evaluate the quality and usefulness of AI assistant responses.
• Review English prompts and responses from an everyday-user perspective.
• Use conversation history from your own eligible personal assistant account.
• Work within a project that organizes contributors into defined age and education groups.

What it pays and takes

This is a remote contract project open to people currently based in the United States. The project runs for 16 weeks, with possible extension opportunities.

• Pay: Paid in USD; the listing does not provide an amount or payment unit.
• Schedule: 20 or more hours per week, up to 40 hours per week.
• Location: Open to candidates currently based in the United States.
• Language: Fluent in English, with comfort evaluating English prompts and responses.
• Account requirement: Actively use Gemini, Instinct, or Muse and have more than two weeks of conversation history in your own personal account.
• Education: Highest completed degree must be a high school diploma, some college, an associate degree, or a bachelor's degree.
• Education limit: Candidates with master's, professional, or doctoral degrees do not qualify.
• Work setup: Able to work remotely.

How it works

Apply on OpenTrain with your resume, then complete the application on the hiring site.

About AI training work

AI training is the human work behind systems that learn from examples, including reviewing and rating AI responses. People with careful judgment help make these systems more useful, accurate, and consistent.